Microchemical Analysis of Rammed Earth Residential Walls Surface in Xiaochikan Village, Guangdong

Abstract

Xiaochikan Village, located in Guangdong Province in South China, is one of the few remaining traditional rammed earth dwellings of the Cantonese ethnic group in the Lingnan region. However, the influence of Zhuhai’s subtropical maritime monsoon climate has led to continuous physical and chemical erosion of the rammed earth walls. For example, cracking occurs due to high temperatures and heavy rain, accelerated weathering occurs due to salt spray deposition, and biological erosion occurs due to high humidity and high temperatures. Therefore, two experimental analysis techniques, X-ray diffraction (XRD) and scanning electron microscopy-energy dispersive spectrometer (SEM-EDS), were used to explore the structural anti-erosion mechanism of the ancient, rammed earth buildings in Xiaochikan Village. The results show that (1) the morphological characteristics of the east and west walls of the rammed earth dwellings in Xiaochikan Village are more similar. The particles on the east wall are regular spherical or polygonal, small, and evenly distributed, while the particles on the west wall are mainly spherical and elliptical, with consistent size and less agglomeration. The surfaces of the particles on both walls are relatively smooth and flat. (2) The core element bases of the four wall samples are consistent, with C, Si, Al, Ca, and Fe as the core, accounting for more than 93%, reflecting the base characteristics of the local alluvial soil “silicate skeleton–carbonate cementation–organic matter residue” and reflecting the “local material” attribute of rammed earth. Except for the south wall sample, the Cl content of the remaining samples exceeds 1%. In the thermal map, Cl shows “pore/interstitial enrichment”, which confirms that the salinization process of marine aerosols with rainwater infiltration and evaporation residue is a common influence of marine climate. (3) The rammed earth walls in Xiaochikan Village consist of three main minerals: Quartz (SiO₂, including alpha-type SiO₂), Calcite (CaCO₃, including synthetic calcite), and Gibbsite (Al(OH)₃).

1. Introduction

1.1. Research Background

As one of the world’s oldest building types, earthen buildings, with their ecological characteristics of locally sourced materials and low carbon footprint, have become a crucial carrier of regional climate adaptation wisdom and local cultural memories [1,2]. Located in the subtropical monsoon climate zone, earthen dwellings in Guangdong’s Pearl River Delta region are subject to long-term environmental stresses such as high temperatures and humidity, typhoons and heavy rains, and microbial erosion [3,4]. The microstructural stability of their wall surfaces (the direct interface between the building and the external environment) directly determines the durability of the buildings and their ability to survive as a heritage site. Xiaochikan Village, a typical traditional Cantonese village in Doumen District, Zhuhai, boasts over 60 extant earthen dwellings dating from the Ming and Qing dynasties (A.D. 1368–1912) to the Republic of China (A.D. 1912–1949) [5]. Their walls utilize a traditional blend of alluvial soil from the Pearl River Estuary, plant ash, and glutinous rice paste. A slurry-coated finish creates a unique, dense protective layer. This layer retains the core advantages of Cantonese earthen architecture: breathability and moisture resistance, while exhibiting distinct micromorphological characteristics distinct from loess cave dwellings in northern China and rammed earth walls in northwest China. It serves as a prime example for studying the surface protective mechanisms of earthen buildings in the humid and hot regions of Lingnan.

The earth dwellings of Xiaochikan Village face dual demands for preservation and utilization as rural revitalization and traditional village preservation strategies advance. On the one hand, field investigations revealed that over half of the village’s earth dwellings have suffered from disrepair, with surface wall deterioration exhibiting microscopic signs of increased porosity, weathering of clay minerals, and microbial attachment, posing a direct threat to structural safety. On the other hand, as important showcases of traditional architectural techniques and intangible cultural heritage, earth dwellings must preserve their historical character and craftsmanship during restoration to avoid the distortion of their cultural value caused by modern material substitution. Against this backdrop, it is crucial to conduct a microstructural analysis of the surface wall structures of the earth dwellings in Xiaochikan Village.

1.2. Literature Review

As a traditional village on the west bank of the Pearl River Estuary, the rammed earth buildings in Xiaochikan Village, Zhuhai, which date back to the Ming and Qing Dynasties, are important physical examples of rammed earth technology in the Lingnan region [6]. Currently, due to environmental factors such as high temperature, high humidity, and typhoons, the rammed earth walls in the village are generally experiencing a decrease in strength, and some buildings are even facing the risk of collapse. In rammed earth buildings, the particles are tightly connected to each other, which can generate greater friction, thereby promoting the stability of the wall structure [7]. Therefore, systematically sorting out the research results related to the microstructure of Lingnan rammed earth has important guiding significance for the protection and restoration of the rammed earth buildings in Xiaochikan Village. Traditional rammed earth in the Lingnan region is mostly based on “three-in-one soil” (local sandy soil, lime, and plant ash) and is often mixed with glutinous rice paste or brown sugar paste as an organic gelling agent [8]. The rammed earth buildings in Xiaochikan Village also follow this process. Oyster shells with a calcium carbonate content of over 90% are also added to some walls to further optimize the structural strength [9,10].

Some existing studies have shown that environmental factors have a multi-mechanism synergistic effect on the erosion of the microstructure of rammed earth dwellings in villages [11,12,13]. Physically, the temperature difference between day and night in the village can reach 15–20 °C and repeated thermal expansion and contraction can easily cause microscopic cracks between rammed earth particles. Chemically, salt spray from the offshore environment penetrates the pores of the wall, generating “salt expansion pressure” through a dissolution-crystallization cycle, causing particles to fall off and gradually increasing the porosity [8,14]. In addition to physical and chemical factors, biological erosion has a particularly significant impact on the north wall. Because the north wall is in a shaded environment for a long time, the humidity often exceeds 85%, providing suitable growth conditions for organisms such as mosses and actinomycetes. The acidic substances secreted by these organisms during metabolism will continuously decompose the mineral components and cementitious materials in the rammed earth, weakening the bonding force between particles and further exacerbating the looseness of the microstructure [15]. Some scholars have found that while the selection of soil for rammed earth is similar in the Mediterranean and China, the different soil compositions and materials constitute the key differences in reinforcing rammed earth [16].

Various experimental analysis techniques are also important ways to explore rammed earth architectural heritage. SEM (Scanning Electron Microscopy), EDS (Energy-Dispersive X-ray Spectroscopy), XRD (X-ray Diffraction), FTIR (Fourier-Transform Infrared Spectroscopy), XRF (X-ray Fluorescence), and other techniques are also very common in architectural heritage [17,18]. For example, infrared thermal imaging, rebound hammer, ultrasonic testing, thermocouple testing, and load tests on the roof frame and floor of Fujian Tulou [19] were used; drip tests on Fujian Tulou sites showed that the degree of erosion was greatest when the rainfall direction was in the range of 30° to 45°, which also provided a reference for the influence of wind and rain on the durability of rammed earth materials [20]. Cui et al. [21] analyzed the development of defects and quality evaluation results of 25 typical rammed earth sites in Gansu Province and conducted a global sensitivity analysis by using fuzzy hierarchical analysis (FAHP), the coefficient of variation method, the entropy method, and the CRITIC method to determine the comprehensive weight of evaluation indicators, thus verifying the applicability of the comprehensive weight-TOPSIS method in the quality evaluation of rammed earth sites. Despite the increasing diversity of analytical methods, the analysis of rammed earth heritage in the Lingnan region of China still requires a long period of exploration.

1.3. Problem Statement and Objectives

While existing research has revealed the fundamental structural characteristics and erosion patterns of rammed earth materials in the Lingnan region, studying the rammed earth structure in the specific area of Xiaochikan Village remains valuable. Future research should focus on rammed earth samples from Xiaochikan Village, conducting targeted analysis using techniques such as SEM-EDS and XRD. This will provide a reference for wall restoration and ultimately contribute to the preservation of traditional rammed earth architecture. This study attempts to explore and answer three main questions: (1) the main architectural structure of the rammed earth dwellings in Xiaochikan Village; (2) revealing the main mineral components of the surface of the rammed earth walls in Xiaochikan Village through XRD; and (3) revealing its main distributed elemental components through SEM-EDS.

2. Materials and Methods

2.1. Study Area: Xiaochikan Village in Guangdong

Xiaochikan Village, the study area, is located northeast of Doumen Town, Doumen District, Zhuhai City, Guangdong Province, China (geographical coordinates: 22°21′–22°24′ N, 113°11′–113°14′ E) [22,23]. It is located on the northwest side of Huangyang Mountain (Figure 1), bordering Huangyang Mountain to the southeast, Shangzhou Village (2.5 km away) to the west, Lianzhou Town Center (5 km away) to the north, and Hutiaomen Waterway (3 km away) to the west. It is a typical “mountain-river-sea” intersection landform with outstanding regional transportation accessibility. It is 55 km away from Zhuhai City, and the Lianzhou Exit of the Guangzhou-Zhuhai West Expressway is 10 km away from the village. It is worth noting that this area has a subtropical maritime monsoon climate, similar to the climate characteristics of Doumen, Zhuhai: the average annual temperature is 22.6 °C, with an extreme high of 38.5 °C and an extreme low of 2.5 °C. The average annual rainfall is 2080.6 mm, with the rainy season running from April to September, accounting for 84.7% of the annual total [24]. There is an average of 11 days of heavy rain per year, with a maximum daily rainfall of 393.7 mm. The average annual relative humidity is 80%, the average annual sunshine hours are 1876.3 h, and the annual accumulated temperature ≥ 10 °C is 8166.6 °C. Typhoons are most frequent from July to September [25], with one to two typhoons directly affecting the area each year and maximum wind speeds of 12–14. The high salt content of the air, with an average annual salt spray concentration of 0.05–0.12 mg/m³, causes continuous physical and chemical erosion of the rammed earth walls. The walls are prone to cracking due to high temperatures and heavy rain, accelerated weathering due to salt spray deposition, and biological erosion due to the high humidity and temperature. Furthermore, due to the presence of Huangyang Mountain (its main peak is 583 m above sea level) to the southeast, the area is less affected by southeasterly winds, with northeasterly and westerly winds being the dominant wind directions (Figure 2).

Xiaochikan Village is one of the best-preserved traditional rammed earth villages in Doumen District, Zhuhai City, with architectural history dating back to the Wanli period of the Ming Dynasty (roughly the late 16th to early 17th centuries) (Figure 3). Protected by relevant policies, Xiaochikan Village is one of the most ecologically well-preserved natural villages in Guangdong Province, offering the unique characteristics of a Lingnan market town known as “Little Guangzhou”. Xiaochikan Village boasts numerous cultural relics and historical sites, including ancient Xiaochikan dwellings, Xiaochikan Stone Street, the Huanggong Ancestral Halls of Youkuan, Zhiwu, Yuting, and the Ma Nanbao Tomb. In 1931, donations from overseas Chinese in Xiaochikan paved a 1500 m-long, 1.6 m-wide granite street running east–west through the village. The pavement is constructed from four slabs, each approximately 0.4 m wide and varying in length. This stone street covers an area of 2400 m² and is the longest and best-preserved stone street in all administrative villages in Zhuhai City. In addition, there are more than 60 well-preserved ancient houses in Xiaochikan Village, all of which are a mixed structure of rammed earth, brick, and wood. They are composed of a gable roof, a dragon boat ridge, a blue brick front wall, a rammed earth gable, and a back wall. They have a distinct Lingnan architectural style from the Qing Dynasty [26,27,28]. Plate-building technology forms the rammed earth walls of the ancient houses in Xiaochikan Village [29]. After hundreds of years, the basic structure has remained intact. The microstructure of the material has a high historical value for studying the ancient residential buildings in the coastal areas of Guangdong. However, due to the typical environment of high temperature and humidity, frequent typhoons, and salt spray in the Pearl River Delta, the rammed earth walls of Xiaochikan Village are facing a serious threat of deterioration.

2.2. Field Investigation and Sample Collection

This study conducted a field survey and sample collection at a typical rammed earth wall building in Xiaochikan Village, Doumen Town, Zhuhai City, from 10:00 AM to 1:00 PM on 29 June 2025. The target was a relatively well-preserved, representative single-unit rammed earth house in the village (Figure 4). According to on-site mapping, the building’s floor plan is approximately 10.90 m × 10.70 m, with a height of approximately 5.87 m from the eaves to the ridge. It has a gabled, double-sloped roof covered with grayish-gray tiles. The exterior walls are original rammed earth, with some traces of later repairs (Figure 5).

The sampling plan adhered to the principles of representativeness, comparability, minimal damage, and traceability and referenced technical specifications for sampling and sample management in the field of earthen site conservation and geotechnical testing. Given the objective of this study to interpret the composition and damage causes of rammed earth, only the surface and near-surface layers of the exterior walls were selected as comparison samples: one sampling point was set on each of the east, south, west, and north exterior walls, for a total of four sites (Figure 5). Each site avoided window openings, post-repair sites, and crack tips and was located in the middle of the wall section, 1.0 m or more from the corners, to minimize boundary effects. The sampling height was controlled at approximately 1.2 m (±0.2 m) above the ground to avoid significant areas of capillary water upwelling and splash erosion near the surface, while also ensuring that subsequent damage observations were not affected. At each site, a 2–3 mm layer of surface weathering/biological attachment was lightly removed with a soft brush. A stainless flat steel shovel was then used to scrape approximately 2–3 cm of near-surface rammed earth from the same location, with a single sample size of 10 g. Each sample is immediately placed in a 30 mL polypropylene sample bottle and sealed tightly. A self-adhesive label is attached to the outside of the bottle, and the location of the sampling point is recorded. At the same time, the point coordinates are marked on the on-site plan and elevation drawings to ensure that indoor testing and field records can be traced in both directions.

2.3. Analytical Techniques

This study combined X-ray diffraction (XRD) with scanning electron microscopy-energy dispersive spectroscopy (SEM-EDS) to analyze the structural corrosion resistance mechanism of the Xiaochikan Village ancient, rammed earth wall under a marine monsoon climate from a three-level perspective: mineral, structure, and elemental analysis. XRD was used to identify and (semi-)quantify the main and secondary mineral phases (such as quartz, calcite, and possible secondary salts), indicating chemical degradation pathways such as dissolution-redeposition, clay interlayer hydration, and salt crystallization through relative content and peak shape changes, and assessing the chemical corrosion resistance potential of the material from the perspective of mineral reactivity and stability. SEM-EDS obtained information such as particle contact relationships, pore throat scale, microcrack density, and elemental enrichment zones of Cl, S, Na, and Ca at the pore edges, characterizing the physical degradation processes such as particle shedding, skeleton disintegration, and salt crystallization expansion and cracking under the action of rain wet–dry cycles and salt spray. The two types of data are interpreted in conjunction at the same sampling point: when the structure exhibits “dense skeleton + low soluble salt + weak edge enrichment”, it is inferred that the structure has high corrosion resistance; conversely, “high carbonate/secondary salt + interconnected pores + significant edge enrichment/microcracks” indicates reduced corrosion resistance. Thus, XRD provides mineralogical evidence of chemical stability and potential reactivity, while SEM-EDS reveals the structural weak points and stress concentration paths. Together, they support the mechanistic interpretation of the corrosion resistance mechanism of rammed earth wall structures and provide quantifiable indicators for protection strategies such as “salt barrier-moisture control-reinforcement”.

2.3.1. X-Ray Diffraction Analysis

Systematic sampling was conducted along the east, south, west, and north directions of the rammed earth building site. Mineral composition analysis was performed using a Rigaku SmartLab high-resolution X-ray diffractometer (Rigaku Holdings Corporation, Tokyo, Japan). The experiment strictly adhered to the X-ray Diffraction Analysis Method for Clay Minerals and Common Non-Clay Minerals in Sedimentary Rocks (SY/T 5163-2018) [30]. Non-destructive testing was performed on bulk samples, each weighing approximately 10 g and precisely controlled to a thickness of 10 mm. Through detailed analysis of the characteristic XRD patterns, the phase composition of clay and non-clay minerals, including quartz, plagioclase, potassium feldspar, calcite, and montmorillonite, was identified. Their crystallinity parameters and semi-quantitative relative content were also determined. This microscopic study, based on mineral crystal field theory, provides core structural data for understanding the weathering mechanisms of rammed earth sites and developing targeted conservation plans.

2.3.2. SEM-EDS Analysis

Scanning electron microscopy-energy dispersive spectroscopy (SEM-EDS) was used to collaboratively characterize the microstructure and elemental composition of rammed earth samples oriented east, south, west, and north. SEM was used to observe differences in particle morphology, particle size distribution, and micropore structure. EDS was used to analyze the elemental composition (e.g., typical rammed earth elements such as Si, Al, Ca, Fe, and Mg) and their regional distribution. This approach further correlated the “structure-composition-erosion resistance” relationship and explored the causes of differences and anisotropic characteristics among samples oriented differently.

This experiment used a Czech TESCAN MIRA LMS (TESCAN, Brno, Czech Republic) as core equipment. The electron microscope’s magnification was flexibly adjusted to 1.00 kX (for macroscopic pore observation), 5.00 kX (for observation of particle aggregates), and 10.00 kX (for observation of microscopic particle morphology), depending on observation requirements. Imaging was performed under a high vacuum environment to ensure clear images of the non-conductive rammed earth samples. EDS analysis was performed with a detection range of C (atomic number 6) to U (atomic number 92) to cover the major and trace elements in the rammed earth. Quantitative calibration was performed using Cu and Al alloy standard samples to ensure relative elemental content errors of less than 5%. The experimental operations sequentially carried out sample pretreatment, fixation and conductive coating, SEM micromorphology observation, EDS elemental analysis, and data processing and statistics. After the sample is fixed, a thin layer of Au conductive coating is applied to improve the imaging quality and signal-to-noise ratio of the non-conductive rammed earth. Then, SEM imaging and EDS analysis are carried out. The coating element signal is not included in the EDS quantification and normalization. The M-series peak of Au is shielded in the spectral fitting to ensure that the interpretation of light elements and halogen elements (such as C, S, and Cl) is not affected by the coating. The samples in four orientations were scanned, and secondary electron images at different magnifications were simultaneously recorded to capture microscopic features such as particle arrangement and pore morphology. EDS analysis was then carried out on typical characteristic areas such as pore edges, particle agglomerates, and abnormal mineral particles in the SEM images, and element distribution spectrum (mapping) were generated to intuitively present the uniformity of the elemental micro distribution. This study used images to qualitatively identify particle size distribution and pore morphology; however, it did not calculate the quantitative particle size distribution and porosity. The elemental data were normalized using the Oxford Instruments software AZtecTEM supporting EDS (https://nano.oxinst.com/products/aztec/aztectem#product-information-tabs, accessed on 4 October 2025), and the average relative contents of key elements such as Ca and Si in the four orientation samples were calculated.

3. Results

3.1. Mineral Composition Comparison Results

By analyzing XRD patterns on the rammed earth wall surface in four directions (east (E), south (S), west (W), and north (N)), we understood the mineralogical composition and stability of the surface material. Figure 6 shows three main minerals: Quartz (SiO₂, including alpha-SiO₂), Calcite (CaCO₃, including synthetic calcite), and Gibbsite (Al(OH)₃). Quartz is a very stable mineral with high hardness and durability, which significantly enhances the overall stability of the material. As can be seen in the figures, Quartz diffraction peaks are prominent in all samples, indicating that they account for a significant proportion of the material in all directions. The environment significantly affects the chemical stability of calcite (including synthetic calcite), with acidic environments being particularly sensitive. Its presence may make the material susceptible to dissolution under acidic conditions, affecting its structural stability. Gibbsite (Al(OH)₃) may decompose under certain temperature and humidity fluctuations, potentially affecting the long-term stability of the material.

The mineral phase groups of the four-directional samples are mainly quartz, calcite, and gibbsite, with similar overall peak shapes; the differences are mainly reflected in the relative peak intensity and peak shape. As shown in Figure 6, the quartz main peak in the east/west-oriented samples is more stable and relatively more prominent, indicating a stronger contribution from the silicate framework; the calcite-related peaks in the south/north-oriented samples are relatively more prominent, suggesting a higher proportion of carbonate cementation and greater sensitivity to acid rain and corrosion environments; gibbsite is identifiable in some samples but is a secondary phase. No typical sulfate/chloride diagnostic peaks were observed in any of the four-directional spectra, indicating that soluble salts exist mostly in trace or subcrystalline forms, and their influence is more evident in the pore edge enrichment and microstructure deterioration of SEM-EDS. The aforementioned mineralogical differences are consistent with the on-site environmental loading: the east/west direction is relatively mildly affected by heat, humidity, and rain, and the quartz-dominated “stable framework” characteristics are more obvious; the south/north direction is under strong sunlight/rain/high humidity conditions, the relative proportion of calcite increases, and it is easy to dissolve and redeposit in wet–dry and acidic environments, reducing the surface corrosion resistance stability.

3.2. Comparison Results of SEM Images

Through SEM imaging analysis of wall rammed earth samples facing east, south, west, and north, it was found that there were significant differences in their micromorphology, particle distribution, and surface characteristics (Figure 7). For other detailed SEM image results at different magnifications, please refer to Appendix A.

In SEM images, the particles from the east wall samples (Figure 7a) are primarily regular in shape, either spherical or polygonal, small in size, evenly distributed with only slight agglomeration, and have a relatively smooth overall surface. This reflects the uniformity and stability of the growth or processing environment for these samples. The particles of the south wall samples (Figure 7b) exhibit significant irregularity, predominantly polygonal, with large variations in particle size, uneven distribution, and significant agglomeration. The surface is rough and accompanied by microcracks, suggesting that the samples in this direction may have been subject to significant stress during their formation.

The particles in the west wall sample (Figure 7c) are mainly nearly round to elliptical, with a medium size range, relatively weak agglomeration, and a relatively flat surface, indicating good particle uniformity and stable and uniform growth or processing conditions. The particles in the north wall samples (Figure 7d) are fragmented or flaky, with significant size variation, a wide distribution, and more prominent agglomeration. The surfaces are also rough, with visible cracks and pores, reflecting strong anisotropy in the microstructure of the samples in this direction, influenced by environmental or processing factors.

Overall, the SEM characteristics of the east and west walls of the rammed earth dwellings in Xiaochikan Village are more similar. The particles on the east wall are regular spherical or polygonal, small, and evenly distributed. The particles on the west wall are predominantly spherical and elliptical, uniform in size, and less agglomerated. The particle surfaces on both walls are relatively smooth and flat. This phenomenon is directly related to the more moderate environmental impacts on east- and west-facing walls under local climatic conditions. As the direction away from direct sunlight and most exposed to concentrated rainfall, these walls are shielded from both the prolonged high temperatures and heavy rainfall from the south and the persistent high humidity from the north. This results in a more intact rammed earth particle structure and a more regular microscopic morphology. However, the microscopic characteristics of the south and north walls differ significantly, and both exhibit signs of structural damage. The south wall exhibits irregular particles, uneven size, and a rough, cracked surface. This feature is due to its position as the surface exposed to intense summer sunlight and the primary rain during the rainy season. The alternating high temperatures, dryness, and wetness, along with rainfall, lead to a breakdown of rammed earth particle agglomerations and surface cracking. The fragmented or flaky particles and pronounced porosity on the north wall indicate that the north direction has the shortest sunlight exposure and relatively high humidity. This long-term humid environment causes the rammed earth particles to flake, leading to the development of a porous structure. These phenomena indicate that the microstructural differences among the four walls of this rammed earth dwelling are essentially due to the differential effects of local natural environmental factors such as sunlight, precipitation, and humidity on the walls facing different directions, ultimately resulting in significant microstructural anisotropy.

3.3. Comparison Results of EDS Element

3.3.1. East Wall Sample

The Ca spectrum signal in the east wall sample is forceful and concentrated, making it the most prominent feature (Figure 8). This is particularly true in the inter-grain cementation region, where the Ca signal appears as continuous bands or flakes, demarcating it from the grain framework (the Si and Al signal regions). The C signal is widely distributed, overlapping with the Ca signal in some areas and scattered across the grain surfaces in others. The signal intensity is moderate and uneven. The Si and Al signals are concentrated in the core of grain aggregates, forming strong, massive signals with clear grain boundaries, distinguishable from the surrounding Ca cement regions. The Fe signal is moderate and evenly distributed at the edges of grain aggregates or within the cement, with occasional strong, dot-like signals at unusual mineral grains. It is speculated that Fe is dispersed within the framework and cement as iron oxides. The Cl signal is slightly enriched at the pore edges, with lower intensity than that of the south wall sample and higher than that of the northwest wall sample. The signal distribution is relatively limited, primarily concentrated near the surface pores, likely due to small amounts of soluble salts remaining after rainwater infiltration.

Furthermore, the east wall’s Ca content, at 30.36% by weight and 15.68% by atomic percentage, is the highest among the four samples (Figure 9,

Table 1. The energy-dispersive spectrometer component analysis results for east wall sample.

Element	Line Type	Wt%	Wt% Sigma	At%
C	K-series	35.1	0.63	60.48
Na	K-series	0.4	0.03	0.36
Mg	K-series	0.14	0.03	0.12
Al	K-series	12.76	0.14	9.78
Si	K-series	14.35	0.15	10.57
S	K-series	0.22	0.03	0.14
Cl	K-series	1.42	0.04	0.83
K	K-series	0.45	0.03	0.24
Ca	K-series	30.36	0.31	15.68
Ti	K-series	0.29	0.05	0.12
Fe	K-series	4.51	0.11	1.68

Source: Author’s statistics.

). This suggests that large amounts of lime or carbonate-rich soil were added during preparation, or that the soil itself is rich in carbonate minerals, resulting in an abundance of cementing material. Theoretically, the rammed earth’s integrity and compressive strength should be higher than those of the other samples. The C content, at 35.10% by weight, likely originates not only from organic matter but also from a significant amount of carbonate carbon (due to the extremely high Ca content, the proportion of carbonate carbon is likely even higher), accounting for over 60% by atomic percentage, indicating that organic carbon and carbonates together form the basis of the high carbon content. The Al content is only slightly higher than that of the north wall sample, with a moderate clay mineral content. The Si content is the lowest among the four samples, with a low proportion of sandy skeletons. Other trace elements indicate moderate impurity content.

3.3.2. South Wall Sample

The spectrum shows continuous, high-intensity, uniform signals for Si and Al in the south wall sample, primarily concentrated in the particle aggregates (Figure 10). At the 50 μm scale, the Si and Al signals almost completely overlap, forming blocky or strip-like areas of a strong signal. These signals correspond to silicate mineral particles within the rammed earth. These particle aggregates form the soil’s skeleton. The continuous signal indicates a relatively uniform distribution of the skeleton particles, with no obvious localized absence or over-concentration, indicating excellent structural integrity. The spectrum signal for Ca is moderate in intensity and primarily distributed in the interstices between particle aggregates, exhibiting a “filling” distribution. For example, at the boundaries between Si-Al framework particles, the Ca signal is significantly enhanced, indicating that Ca fills the interstices between the framework particles as a cement, exerting a cementing effect. The relatively continuous Ca signal distribution indicates that the cement fills the interstices uniformly, without the presence of excessive porosity caused by localized absence of cement. The spectrum signal for C is high in intensity but unevenly distributed, displaying a “scattered dot pattern with localized banding”. Scattered point signals may correspond to tiny plant debris particles, while localized band signals may represent the remnants of fibrous organic matter. This uneven distribution is consistent with the natural retention of organic matter in rammed earth. Furthermore, the C signal is not concentrated in specific areas (such as pore edges), indicating that organic matter has not undergone significant migration due to environmental factors. The Fe spectrum signal intensity is moderate and relatively evenly distributed, occurring in both skeletal particles and cementing areas, with occasional localized point signal enhancements. This distribution indicates that Fe is uniformly dispersed in the rammed earth matrix as a trace impurity, with no significant enrichment or loss. The auxiliary cementation effect of iron oxides is evident throughout the sample.

C, at 40.77% by weight and 64.70% by atomic percentage, is the most abundant element in the sample (Figure 11,

Table 2. The energy-dispersive spectrometer component analysis results for south wall sample.

Element	Line Type	Wt%	Wt % Sigma	At%
C	K-series	40.77	0.74	64.7
Na	K-series	0.07	0.03	0.06
Mg	K-series	0.09	0.03	0.07
Al	K-series	13.81	0.19	9.75
Si	K-series	23.39	0.31	15.87
S	K-series	0.31	0.04	0.18
Cl	K-series	0.09	0.03	0.05
K	K-series	0.48	0.04	0.23
Ca	K-series	14.15	0.2	6.73
Ti	K-series	0.3	0.06	0.12
Fe	K-series	6.56	0.16	2.24

Source: Author’s statistics.

). This high C content may be due to two factors: first, residual organic matter such as plant debris and fibers mixed into the rammed earth during preparation; second, carbonaceous components in the soil itself or carbon deposited in the environment later. From an atomic perspective, C accounts for over 60% of the atoms in the microstructure, making it a key element influencing the organic matter content of rammed earth. Si, at 23.39% by weight and 15.87% by atomic percentage, is the second most abundant element. Si is a core component of silicate minerals, forming the “skeleton” of rammed earth and fundamentally determining its mechanical strength. Its high content indicates a high proportion of sandy or silty minerals in the sample, resulting in a more stable skeleton structure. Ca, primarily derived from carbonate minerals (such as calcite CaCO₃) in the rammed earth or from lime added during preparation (which forms a calcium carbonate cement upon hydration), is a crucial “binder” for rammed earth, filling the gaps between mineral particles, strengthening the bonds between particles, and enhancing the integrity of the rammed earth.

3.3.3. West Wall Sample

The Si and Al spectrum signals in the west wall sample are forceful, dominating the image and exhibiting a “large, continuous, blocky” distribution that covers most of the sample’s microscopic area (Figure 12). The Si and Al signals completely overlap, forming densely distributed, large-scale framework aggregates (mostly larger than 20 μm) with narrow inter-aggregate gaps. This indicates a high content of sandy and clay mineral particles, a dense framework structure, and the absence of significant macropores or framework vacancies. This distribution perfectly matches the high Si and Al content and is typical of sandy rammed earth microstructures. The Fe signal is moderate in intensity and relatively uniformly distributed, primarily accompanying the Si and Al signals within framework aggregates, with occasional localized, dotted areas of enhancement. Due to the densely packed framework particles, the Fe distribution is widespread, helping to strengthen the internal bonding of the framework particles. The signal intensity of Cl is moderate, and its distribution characteristic is “framework interstitial enrichment”—it is mainly concentrated in the narrow gaps between Si-Al framework aggregates. Because the gaps are narrow and there is no large amount of Ca cement filling, Cl-containing water is retained here and evaporates and remains, forming interstitial enrichment. However, due to the narrow gaps, the enrichment area is distributed linearly, rather than being enriched in sheets at the pore edges, like in the north wall sample. The result shows that although there are salt residues in this sample, the pore space is limited, and the salt enrichment range is relatively limited.

Si, at 38.03% by weight and 28.03% by atom, is the highest of the four samples and a core characteristic element (Figure 13,

Table 3. The energy-dispersive spectrometer component analysis results for west wall sample.

Element	Line Type	wt%	Wt % Sigma	At%
C	K-series	29.53	0.58	50.89
Na	K-series	0.38	0.02	0.34
Mg	K-series	0.13	0.02	0.11
Al	K-series	18.79	0.16	14.42
Si	K-series	38.03	0.32	28.03
S	K-series	0.48	0.02	0.31
Cl	K-series	1.33	0.03	0.77
K	K-series	0.55	0.02	0.29
Ca	K-series	5.63	0.06	2.91
Ti	K-series	0.24	0.03	0.1
Fe	K-series	4.92	0.08	1.83

Source: Author’s statistics.

). This high Si content indicates a high proportion of sandy silicate minerals in this sample, a typical “sandy skeleton-dominated” rammed earth, with a strong mechanical stability and deformation resistance. Al, at 18.79% by weight and 14.42% by atom, is also the highest of the four samples and, together with Si, forms the basis for the high silicate mineral content. This high Al content indicates a rich content of clay or feldspar minerals. Clay minerals enhance the plasticity of rammed earth, making it easier for sandy particles to agglomerate, while feldspar minerals strengthen the skeleton. The carbon content, at 29.53% by weight and 50.89% by atom, is the lowest among the four samples, indicating low organic matter and carbonated carbon content. These findings may be due to the wall’s more intense weathering or leaching, which resulted in organic carbon loss. Furthermore, the cement content is low, resulting in a low proportion of carbonate. The calcium content, at 5.63% by weight and 2.91% by atom, is the lowest among the four samples, indicating minimal carbonate cement or lime addition, indicating a lack of cementing material. The integrity of the rammed earth relies primarily on the viscosity of clay minerals and interparticle friction, rather than chemical cementation.

3.3.4. North Wall Sample

The carbon spectrum signal in the north wall sample is high and evenly distributed (Figure 14). Unlike the uneven distribution in the south wall sample, the carbon signal in this sample is diffuse, with no distinct concentrated areas, such as dots or stripes. It is speculated that the organic matter may have undergone a certain degree of decomposition or dispersion and was evenly mixed in the rammed earth matrix in the form of finer particles, without any local residual accumulation. The Si signal intensity is high and concentrated in the particle agglomerate area, forming a blocky strong signal area. However, the size of the agglomerates is slightly smaller than the Si-Al agglomerates in the south wall sample, and the gaps between the agglomerates are wider, corresponding to the slightly lower Si content. The skeleton particles are still evenly distributed without obvious vacancies. The Ca signal intensity is high, and it is the strongest signal in the thermodynamic diagram after C and Si. It has a wide distribution range and fills the gaps between Si aggregates and slightly overlaps with the Si signal in some areas. This indicates that the Ca content is sufficient. The cement not only fills the gaps but also may form a thin cementing layer on the surface of the particles, enhancing the bonding force between the particles and the cement. The signal is continuous without obvious breakpoints, and the cementation uniformity is excellent. The Cl heatmap signal intensity is moderate, making it the clearest among the low-content elements. Its distribution is notable—it is primarily concentrated at pore edges and on the outer surfaces of particle aggregates, exhibiting an “edge enrichment” phenomenon. At a 50 μm scale, the Cl signal around pores is significantly stronger than in the matrix. This attribute is due to Zhuhai’s high rainfall and humidity. When Cl-containing rainwater or moisture penetrates the rammed earth, soluble chloride salts are left behind in the pores due to evaporation, resulting in edge enrichment. This characteristic is consistent with the salinization trend caused by the maritime climate.

C, at 39.98% by weight and 64.35% by atom, is second only to the south wall sample and remains the most abundant element in the sample (Figure 15,

Table 4. The energy-dispersive spectrometer component analysis results for north wall sample.

Element	Line Type	Wt%	Wt % Sigma	At%
C	K-series	39.98	0.63	64.35
Na	K-series	0.37	0.03	0.31
Mg	K-series	0.37	0.03	0.29
Al	K-series	9.65	0.11	6.92
Si	K-series	20.4	0.22	14.04
S	K-series	2.77	0.05	1.67
Cl	K-series	1.69	0.04	0.92
K	K-series	1.13	0.04	0.56
Ca	K-series	20.19	0.23	9.74
Ti	K-series	0.14	0.04	0.06
Fe	K-series	3.30	0.1	1.15

Source: Author’s statistics.

). The C origin is consistent with that of the south wall, primarily organic and carbonaceous components, accounting for over 60% of the total atomic content. This value indicates that this sample also contains a high organic matter content, similar to that of the south wall. Si, at 20.40% by weight and 14.04% by atom, is the second most abundant element, slightly lower than the Si content in the south wall sample. Ca, at 20.19% by weight and 9.74% by atom, is significantly higher than the Ca content in the south wall sample (14.15%) and is a prominent core element characteristic of this sample. The high Ca content indicates a higher amount of carbonate cement or lime added to this sample, resulting in a more robust cementing base and potentially greater integrity. S, at 2.77% by weight and 1.67% by atom, is the highest of the four samples and the most prominent secondary component characteristic of this sample. The high sulfur content may be due to two factors: first, the high content of sulfate minerals in the local soil; and second, Zhuhai’s proximity to the ocean. Sulfur compounds in the atmosphere (such as sulfate in marine aerosols) settle and penetrate the rammed earth, leading to sulfur enrichment. The Cl content, at 1.69% by weight and 0.92% by atomic percentage, is also the highest among the four samples, significantly higher than the 0.09% found in the south wall sample. Cl is a typical soluble salt element, and its high Cl content is closely related to Zhuhai’s maritime climate. Sodium chloride (sea salt) in the marine atmosphere penetrates the rammed earth with rainwater or moisture, forming soluble salt residues.

Overall, the core element composition of the four samples is consistent, with C, Si, Al, Ca, and Fe constituting over 93% of the total. This reflects the fundamental characteristics of the local alluvial soil, described as “silicate skeleton-carbonate cementation-organic matter residue” [32], and highlights the “local material” attribute of rammed earth. Except for the south wall sample, the Cl content of the remaining samples exceeds 1%. In the thermal map, Cl shows “pore/interstitial enrichment”, confirming that the salinization process of marine aerosols with rainwater infiltration and evaporation residues is a common influence of marine climate.

As shown in

Table 5. Elemental composition from EDX analysis on the sample from the walls.

Wall	East		South		West		North
Metric	wt%	At%	wt%	At%	wt%	At%	wt%	At%
C	35.1	60.48	40.77	64.7	29.53	50.89	39.98	64.35
Na	0.4	0.36	0.07	0.06	0.38	0.34	0.37	0.31
Mg	0.14	0.12	0.09	0.07	0.13	0.11	0.37	0.29
Al	12.76	9.78	13.81	9.75	18.79	14.42	9.65	6.92
Si	14.35	10.57	23.39	15.87	38.03	28.03	20.4	14.04
S	0.22	0.14	0.31	0.18	0.48	0.31	2.77	1.67
Cl	1.42	0.83	0.09	0.05	1.33	0.77	1.69	0.92
K	0.45	0.24	0.48	0.23	0.55	0.29	1.13	0.56
Ca	30.36	15.68	14.15	6.73	5.63	2.91	20.19	9.74
Ti	0.29	0.12	0.3	0.12	0.24	0.1	0.14	0.06
Fe	4.51	1.68	6.56	2.24	4.92	1.83	3.3	1.15

Source: Author’s statistics.

, the elemental composition of the samples from the four walls was mainly C, Si, Al, Ca, and Fe, all accounting for >93 wt% (East 97.08%, South 98.68%, West 96.90%, North 93.52%), but there were significant directional differences in the framework and cementation type. The west wall had significantly higher Si and Al content (Si 38.03 wt%, Al 18.79 wt%, Si/Al ≈ 2.0, Ca only 5.63 wt%, Ca/Si ≈ 0.15), indicating a robust framework mainly composed of quartz/aluminosilicate. The east wall had a prominent Ca content (30.36 wt%, Ca/Si ≈ 2.12), which better reflected the carbonate cementation characteristics. The south wall had higher Si and Al content but lower Ca content (Si 23.39 wt%, Al 13.81 wt%, Ca 14.15 wt%, Ca/Si ≈ 0.60), suggesting that carbonate leaching may have occurred under strong sunlight and rain. The north wall exhibits a mixed matrix (Si 20.40 wt%, Ca 20.19 wt%, Ca/Si ≈ 1.0). Regarding soluble salt indicators, Cl ≥ 1 wt% is observed in all directions except the south wall (East 1.42, West 1.33, North 1.69 wt%), with higher S on the north wall (2.77 wt%), consistent with salt accumulation and crystallization stress under high humidity/low sunlight conditions. The south wall has only 0.09 wt% Cl and 0.31 wt% S, consistent with salt erosion by rain. C content is generally high (~29.53–40.77 wt%, atomic fraction approximately 50.89–64.70%), appearing alongside higher Ca content on the east/south/north walls, possibly indicating the presence of carbonates or organic residues. In summary, the west wall is predominantly composed of a siliceous framework and exhibits low chemical solubility sensitivity. The east wall has a high proportion of carbonate cementation and is more sensitive to acid rain, dissolution, and redeposition; the north wall has the highest salt load and salt swelling risk. The south wall has the lowest salt content but may suffer from surface weakening due to carbonate loss.

4. Discussion

At this stage, this study can only provide evidence of the role of soluble salts based on the semi-quantitative and spatial distribution of S and Cl in EDS and correlate it with visible deterioration phenomena such as salt frost/dust on the wall surface so as to avoid misinterpreting the elemental signals as a definitive identification of minerals. Through on-site environmental surveys, XRD analysis, and SEM-EDS analysis, the long-term stability of the rammed earth walls of Xiaochikan Village, under the subtropical maritime monsoon climate, was revealed. The study revealed that the microstructure (particle morphology, porosity, and bond strength) and chemical composition (element distribution and mineral content) of the four walls exhibit significant directional variations due to their exposure to different environments. The east wall maintains its structural density thanks to high carbonate cementation, while the south wall is susceptible to cracking due to high temperatures and heavy rain. Salt spray deposition accelerates weathering of the west wall, while the high humidity triggers biological erosion of the north wall. Based on the “environment-microstructure-macrodegradation” linkage mechanism and combining traditional rammed earth craftsmanship with modern conservation techniques, a conservation strategy tailored to each orientation was proposed to support the living legacy of the village’s rammed earth architecture.

4.1. East Wall: Potential Impacts of Carbonate Balance and Diurnal Temperature Variation

The East Wall is the most structurally stable of the four rammed earth walls in Xiaochikan Village. Its characteristics are closely related to the regional environment and material composition. Microstructurally, the rammed earth particles in the East Wall are mostly regular, spherical, or polygonal, small, and evenly distributed, with only slight agglomeration. The overall surface is relatively smooth, and the porosity is well within a reasonable range. This morphological characteristic is directly related to the local east-facing walls being more moderately exposed to the natural environment. They are shielded from the direct impact of prolonged high temperatures and heavy rains from the south and mitigate the persistent effects of long-term high humidity from the north, thus preserving the intact rammed earth particle structure. Conservation strategies for the East Wall should focus on stabilizing the carbonate balance and strengthening the cementation strength. Based on the traditional Lingnan rammed earth craft, it is recommended that a mixed coating of “oyster shell powder + glutinous rice paste” be used for protection in the future. The calcium carbonate content in oyster shell powder exceeds 90%, which can artificially supplement the carbonate in the wall and reduce natural dissolution loss. The glutinous rice paste can enhance the adhesion between the coating and the wall and prevent the coating from falling off. In addition, this formula is in line with the process logic of the traditional rammed earth “three-in-one soil + organic gelling agent” of Xiaochikan Village, ensuring that the protection measures do not damage the historical appearance of the wall [9,10].

4.2. South Wall: Synergistic Impacts of High Temperature and Heavy Rainfall

The South Wall is one of the most significantly stressed walls in Xiaochikan Village. Its characteristics are primarily due to the synergistic effects of high temperature and heavy rainfall. Xiaochikan Village has a subtropical maritime monsoon climate, with an average annual temperature of 22.6 °C (maximum 38.5 °C) and an average annual sunshine duration of 1876.3 h. Intense summer sunshine directly impacts the South Wall. Furthermore, rainfall during the rainy season (April–September) accounts for 84.7% of the annual total, with a maximum daily rainfall of 393.7 mm. Heavy rainfall exerts a significant direct erosion effect on the South Wall. Microstructurally, the rammed earth particles in the South Wall are irregular polygons with a wide range of particle sizes (3–20 μm in diameter) and uneven distribution. Significant agglomeration is evident, and the surface is rough with fine cracks, reflecting the significant environmental stresses sustained during the formation and use of this wall. The protection strategy for the south wall should focus on solving the problems of crack expansion and surface peeling caused by high temperature exposure and heavy rain while strengthening the bonding strength. In the future, it is recommended that a removable bamboo sunshade (30 cm spacing) be built along the outer side of the south wall at the physical protection level to maintain the ventilation of the wall while reducing the surface temperature by 5–8 °C and reducing the thermal expansion and contraction of particles caused by high temperature. Traditional gray tile eaves (80 cm wide) should be added to the top of the wall to replicate the “rainproof eaves” of Lingnan folk houses [33,34,35], guiding rainwater away from the wall and preventing heavy rain from directly eroding the surface fine particles [36].

4.3. West Wall: Weathering Effects of Salt Fog Enrichment and Insufficient Cementation

The characteristics of the West Wall are determined by salt fog erosion and insufficient cementation of the sandy skeleton, which is closely related to the location environment of Xiaochikan Village. From a macro perspective, the village is in the alluvial plain on the west bank of the Pearl River Estuary. There is no high terrain blocking the west side. The salt fog in the marine climate comes not only from the southeast monsoon but also from the local circulation of the Hutiaomen waterway on the west side (such as the southwest monsoon and westerly wind). Although Huangyang Mountain is in the southeast (the main peak is 583 m above sea level), it can weaken the offshore salt fog carried by the southeast monsoon, but it cannot block the “near-source salt fog” formed by evaporation from the nearby waterway on the west side. In addition, the West Wall is close to the village water system, and the groundwater level is high, which further aggravates the upward migration of salt, superimposed on the influence of salt fog, forming a “double salt erosion” effect [37]. Therefore, it is highly recommended that the protection strategy of the West Wall should focus on “salt blocking, salt drainage, and salt swelling prevention”, considering the stability of the sandy skeleton and the prevention and control of salt fog erosion. In terms of water repellency and salt resistance, traditional tung oil can be used to enhance the water repellency of the wall surface [38] and reduce salt adhesion. This material is in line with the protection principles of Xiaochikan Village and avoids the impact of chemical reagents on the environment. In terms of drainage and moisture-proof optimization, a blue brick moisture-proof layer (50 cm high) is added at the base of the west wall to match the village’s rammed earth and brick-wood mixed building structure to block salt migration caused by rising groundwater. A 10 cm wide and 15 cm deep gravel drainage ditch is excavated along the foot of the wall, and a gravel layer is laid to accelerate the drainage of rainwater and reduce the retention time of salt on the wall [39].

4.4. North Wall: Synergistic Destruction by High Humidity and Biological Activity

The synergistic effect of high humidity and biological erosion determines the characteristics of the north wall. It is the wall with the most significant loose structure in Xiaochikan Village. The north side of Xiaochikan Village has the shortest sunshine hours, and the north side accounts for the lowest proportion in the annual average sunshine hours of 1876.3 h. As a result, the north wall is in a shaded environment for a long time, and the environmental humidity often exceeds 80%, providing suitable growth conditions for organisms such as mosses and actinomycetes [40,41]. Therefore, it is recommended that the protection strategy of the north wall should focus on “humidity control-bacterial inhibition-loose structure repair”, considering both biological erosion prevention and control and wall permeability. In terms of antibacterial protection, TiO₂/chitosan nanocomposite coating is used, which has an antibacterial rate of 92.19% against Staphylococcus aureus [42]. Chitosan is derived from shells (which is consistent with the oyster shell component in the rammed earth of Xiaochikan Village), which is in line with the tradition of using local materials. In terms of humidity control, moisture-resistant trees (such as local banyan trees) are planted around the north wall to reduce ground evaporation and lower the local environmental humidity through vegetation transpiration; ventilation holes are set at the bottom of the wall (with a spacing of 1 m) to promote air circulation inside and outside the wall and avoid water vapor retention. At the same time, the ventilation holes are covered with wooden grilles to coordinate with the traditional architectural style [43].

5. Conclusions

Xiaochikan Village, Doumen Town, Zhuhai City is in the subtropical maritime monsoon climate zone. Long-term exposure to high temperatures, high humidity, salt spray, and heavy rain has put the local traditional rammed earth buildings at risk of accelerated deterioration. This study aimed to reveal the compositional and microstructural differences and dominant failure mechanisms of the exterior walls of typical rammed earth dwellings in Xiaochikan Village along different orientations and to propose differentiated conservation strategies. Methodologically, targeted sampling of the exterior walls was combined with SEM-EDS and XRD to systematically characterize their morphology, elemental composition and spatial distribution, and mineralogical phases. The following conclusions were drawn:

(1)

There are significant differences in orientation between the four walls. The east wall is characterized by high carbonate cementation (Ca = 30.36 wt%), with regular and dense particles. The west wall is dominated by a silicate skeleton, with the highest Si and Al contents (Si = 38.03 wt%, Al = 18.79 wt%) and the lowest Ca, indicating insufficient cementation and domination by particle friction and cohesion. The north wall shows marginal enrichment of soluble salt indicator elements (Cl = 1.69 wt%, S = 2.77 wt%), suggesting salinization risk. The south wall has irregular particles, fine cracks, and obvious agglomeration damage, reflecting the synergistic damage caused by strong sunshine, wind-driven rain, and rainstorm erosion. The above differences are consistent with the local dominant wind direction, rainfall, and sunshine exposure conditions.

(2)

Mineral assembling and engineering characteristics are obvious. The four-direction samples generally contain quartz, calcite, hydrous aluminum silicates (kaolinite group), and aluminum hydroxide (gibbsite). Quartz provides framework stability; calcite is dissolved in acidic/wet environments; layered aluminum silicates have water absorption and expansion properties, and the presence of aluminum hydroxide may also enhance the volume change effect during the wet–dry cycle, jointly triggering water absorption-swelling-contraction-surface erosion.

(3)

Mechanism chain and vulnerability pattern. Wind-driven rain, salt fog deposition, and high humidity in the marine climate drive carbonate dissolution, salt crystallization, and biological attachment. These effects dominate in different directions, forming a vulnerability pattern where the east wall is relatively stable, the south wall is prone to cracking, the west wall is prone to erosion, and the north wall is prone to tidal salt. This pattern provides a basis for proposing a directional protection and restoration method for rammed earth walls.

This study also has some shortcomings: (1) Due to constraints related to powder sample collection and heritage conservation, this paper does not conduct quantitative analysis of particle size and porosity based on polished cross-sections/3D stereoscopic views. Currently, assessments of homogeneity and agglomeration are only qualitative, based on images. In the future, when conditions permit, limited two-dimensional image measurements (equivalent circle diameter, aspect ratio, roundness, and pore throat dimensions, n ≥ 200/sample) will be conducted based on acquired SEM images. The two-dimensional approximations and representative boundaries will be clarified to provide semi-quantitative support for morphological description without introducing additional damage to the material. We understand this is a certain difficulty for sampling traditional village residential heritage, and we can still work in this direction in the future. (2) To control length and minimize sampling damage, this paper did not conduct qualitative and quantitative analysis of the mineral phases of weathering products: stoichiometry of water-soluble ions (Cl⁻, SO₄²⁻), phase identification of trace secondary salts, and semi-quantitative XRD full spectrum fitting of trace phases are still lacking. Existing SEM-EDS only provides semi-quantitative evidence and pore edge enrichment patterns at the elemental level and cannot replace mineral-level confirmation. Future work will focus on: (a) Quantitative analysis of water-soluble salts and conductivity/pH measurement to calculate total salt load and ion ratio. (b) Phase confirmation of typical salt crystals using Raman/FTIR. (c) Semi-quantitative analysis of secondary salts using XRD-Rietveld. (d) Establishing a correlation model between salt content indicators and macroscopic deterioration grading and micropore structure parameters (equivalent pore throat, porosity) to improve the discriminative power and comparability of rammed earth durability assessment. (3) The time dimension is inadequate. One-time summer sampling cannot reflect the process influence of seasonal cycles (typhoon season and dry season), and there is also a lack of rapid retesting and long-term tracking after heavy rain. The next step will be to expand the grid sampling of multiple buildings and multiple depths, superimpose in situ moisture or salt monitoring and mechanical-multi-field coupling verification, and carry out field tests and evaluations of orientation-oriented materials for salt blocking, moisture control, and cementation repair to form a scalable Lingnan coastal rammed earth protection technology system.